nyame akuma issue 032

7/26/2019 Nyame Akuma Issue 032

1/68

NYAME AKUMA

No. 32 December 1989

The first issue of Nyame Akuma,

published in October

1972

was edited by

Peter Shinnie, who, having served as

midwife (if not mother), continued to

nurture the periodical for a full decade, at

which time he relinquished editorial

responsibility to David Lubell. Having just

stepped into David's shoes, want to

acknowledge the dedication and effort of

my predecessors, as well as the many

people who have assisted them and the

myriad of contributors. The fact that

Nyame Akuma has recently celebrated its

seventeenth birthday is, of course, a

testimonial to its unique place in the

periodical literature on African

archaeology. But it is no less a tribute to

the industry and commitment of those who

have labored to create and sustain it. On

behalf of everyone who benefits from the

existence of Nyame Akuma, warmest

thanks to my predecessors and those who

worked with them.

Uncomfortable though it is to enter a

parade out of step, I am afraid that this

issue of Nyame Akuma will reach its

subscribers a bit late. For that, my

apologies; the reasons include my

inexperience in editing this periodical and

a nine-month delay in my obtaining the

institutional support mentioned in David

Lubell's last editorial. 1 shall try to

adhere more closely to the publication

schedule in future issues.

It will help if contributors would

prepare their manuscripts with the same

care and attention to detail that would be

expected in a submission to a refereed

journal. Although the Instructions for

Contributors (opposite) are imprecise and

editorial standards for Nyame Akuma

have traditionally been somewhat elastic,

certain generally held, minimal criteria for

manuscript preparation really should be

honored. For example, all hard copy

versions of manuscripts, including those

accompanied by disks, should be double-

spaced original typescripts or copies of

comparable quality on

8 3

x

11 (215

mrn x

278 mm) sheets. In addition, care should be

taken to

insure that all citations in text

actually appear in the list of references and

that the form of the bibliography

corresponds to the style evident in recent

issues of Nyame Akuma, a slightly

modified version of the bibliographic style

for American Antiquity. Speaking of style,

I would like to take this opportunity to

solicit suggestions for improving either the

appearance or the content of Nyame

Akuma.

As regards content, let me underscore

the breadth of our net; we really do

welcome contributions on all aspects of

African archaeology, with the exception

of articles on Pharaonic Egypt or Classical

North Africa. The exclusion of historical

material from North Africa, for which

there are numerous publication outlets, does

not extend to sub-Saharan Africa. Thus,

contributions on the historic period are,

indeed, welcome from any part of sub-

Saharan Africa, as is evident in the

contributions from, for example, Tanzania

in this issue.

I plan to attend the biennial SAfA

meetings in Gainesville (see under

Meetings ) and will be happy to hear

comments, criticisms, etc. from anyone

present at the conference.

John Bower


2/68

NYAME AKUMA

No 32

December

1989

Predynastic Settlement

Patterns in the Hiw-

Semaineh Region

Upper Egypt

Kathryn Bard

Department of Archaeology

Boston University

675 Commonwealth Avenue

Boston MA 02215 USA

An expedition to locate and investigate

the status of Predynastic sites in the Hiw-

Semaineh region was conducted in May-

June, 1989, by the au thor of this

contribution. The project was planned and

coordinated with the assistance of Farouk

El-Baz, Director of Boston University's

Center for Remote Sensing, and Fekri A.

Hassan, Washington State University.

Funds for the project were provided by the

National Geographic Society.

Field work consisted of a reconnaissance

archeological survey, systematic survey of

selected areas, geoarcheological investi-

gations, and a study of the setting and

characteristics of two sites. Field work

began by relocating the archeological sites

mentioned by Sir Flinders Petrie (1901).

Cemeteries of Abadiyeh and Hiw

The Hiw-Semaineh region was the

focus of field work by Sir Flinders Petrie in

1898-99. In one field season Petrie

excavated five Predynastic cemeteries,

none cemeteries dating to pharaonic times,

and two Roman-period cemeteries along a

16 km strip from Hiw to Semaineh (Petrie

1901). Petrie also recorded evidence of

Predynastic villages between Abadiyeh

and Semaineh, but he did not excavate

them. Investigations of a 1986 Landsat

Thematic mapper image of the Nag

Hammadi-Qena region indicated that

much of the area above the floodplain from

Abadiyeh to Semaineh was not presently

under cultivation, and Predynastic sites

might still be preserved there.

Reconnaissance in 1989 showed that

several Predynastic sites of Petrie's had

been destroyed. Cemetery U east of the

large Muslim cemetery of Hu, is now the

site of a gravel quarry. Cemetery

R

was

located in the housing area for the Mag

Hamrnadi Aluminum Factory, and has been

destroyed by bulldozing.

At Abidiyeh we relocated the remains

of a Predynastic village which Petrie

recorded as entirely plundered. Situated

on a spur south of the floodplain, most of

the area encompassing this Predynastic

village was cultivated from 1955 to 1965.

Part of the village to the southeast was

recently bulldozed. This site was named

HG after the modern village of Halfieh

Gibli, on the western edge of the

Predynastic site (Fig. 1).

Halfieh Gibli

From the northern edge of the spur on

which the Halfieh Gibli (HG) site is

situated, the site extends south for 180 m to

the edge of presently cultivated land. The

site was gridded and a 20% surface

collection was randomly selected and

collected. Tests pits of 1 x 1 m were

excavated in 10% of the grid units.

Site HG was covered with the sherds of

typical Predynastic wares, black-topped

red ware (BTRW) and rough-ware (R-

ware), but no architecture was visible on

the surface. The area of the heaviest

concentration of artifacts was in the

southeastern quadrant of the site where

four large limestone blocks located.

Although much of the site had been

previously disturbed by cultivation, a test

pit in the area of the four limestone blocks

was done in a stratified midden. Charcoal

samples taken from test pits are now being

processed for radiocarbon dating.

Lithic artifacts collected from site HG

include retouched flakes and blades, and

several bifacial blades. A few scrapers

were collected and several fragments of


3/68

NYAME

A K W A

No 32 December 1989

sickle blades were found in test pit

deposits. No projectile points were found in

the sieved test pit deposits, but a large

stone for grinding

grain

was collected on the

surface. Other stone tools include pounding

and polishing stones for stoneworking,

possible for the stone vessels that have

een found in elite Predynastic graves.

Abadiyeh

Cemeteries

East of site HG we relocated Petrie's

Cemetery A, which he dated to the Old

Kingdom. The mastaba which he described

there has now been partially eroded away

on the north slope of the spur. Petrie's

Predynastic Cemeteries and C extend to

the south and east of Cemetery A and

appear to have een thoroughly excavated

by him.

At the eastern end of Abadiyeh we

relocated a stretch about km long where

Petrie noted Predynastic villages (his F

sites). This area is now covered with

houses and threshing floors, and

Predynastic sherdsoccur on the surface.

Semaineh Cemeteries

At the eastern end of Petrie's survey

near the modern village of Semaineh, we

relocated Petrie's Cemetery H. which

dates mainly to the Terminal Predynastic

period or Nagada

111

On

the spur to the

north and east of Cemetery H are the

remains of a Predynastic village not

mentioned by Petrie, which we called site

SH

(Fig.

1 . Site SH extends about 80 m

across the northern end of the spur and for

about 50 60 m north-south.

Site SH was gridded and a

10

surface

collection was randomly selected and

collected. Typical Predynastic sherds of

TRW

were found here but in much smaller

quantities than at site HG. The

predominant ware collected on the surface

of SH was R-ware. A few rims of Meydum

ware were also collected on the surface,

suggesting an Old Kingdom component at

the site as well. Test pits of x m

indicated that deposits of the Predynastic

village are thin. However, in the northern

end of the site, where fragments of mud-

bricks were found on the surface, the


4/68

NY ME KUM

No

32

December 1989

remains of mud-brick were seen in a test pit

below 10an

At site SH there was evidence of ground

stoneworking, as at site HG. Tools for

stoneworking included polishers and

multifaceted grinding stones of quartzite

and a dark red igneous rock, perhaps a

metavolcanic from the Wadi Hammamat.

The polishing stones are too large to have

been used to polish pots. Analysis of thin

sections from several of these tools now

being done. Hammerstones and large chips

from the stoneworking were also found. It

can perhaps be suggested from the surface

evidence that the Hiw region was a center

for stone vessel manufacture in Predynastic

times and the raw materials for these

vessels were imported from the Wadi

Hammama .

Because of its proximity to the

Terminal Predynastic cemetery H and its

thin deposits, which are indicative of a

more short-term occupation, site SH almost

certainly dates to this period. Charcoal

samples from site SH are currently being

processed for radiocarbon dating, and it is

hoped that dates from these samples will

help resolve the controversy of the

beginning of the First Dynasty and the

starting date for the Egyptian chronology.

Although there is an established

Predynastic sequence of artifact types b sed

on burials Petrie 1901; Kaiser 1957; Kemp

19821, the wares and relative quantities of

these

from the two Predynastic settlement

sites at Abadiyeh and Semaineh were

different from those known for Predynastic

burials. From our surface collection, BTRW

was used at settlements throughout the

Predynastic sequence, whereas in burials

this ware disappears by the Terminal

Predynastic. R-ware was the most common

ware found at sites HG and SH, and

included some sherds 2.0 m or more thick,

which may have come from large storage

units. Decorated-ware sherds were not

found at the two settlement sites, and this

ware may have been primarily used for

Predynastic grave goods. No burnished

Badarian ware was found at either site,

indicating a later date of occupation in the

Predynastic sequence.

Test pits indicated that two loci within

site SH

may

have been threshing floors, but

the site was not previously excavated or

cultivated. There is some evidence of

stratified remains at both sites HG and SH,

which is extremely rare for Predynastic

settlements. It is hoped that site SH,

which is of Terminal Predynastic date,

will provide missing information for the

period when the Early Dynastic state was

formed in Egypt. Both of these sites are

threatened by irrigation projects which

will reclaim 17,000 acres of land. It is

imperative that the sites excavated as soon

as possible before the information and the

sites are lost forever.

eferences

Kaiser von Werner

1956

Stand und Probleme der agyptischen

Vorgeschichtsforschung.

Zeitschrift

fur agyp t i schc Sprache und

Altertrrmskundc

81: 87-109.

Kemp

Barry

J

1982

Automatic analysis of Pre-dynastic

cemeteries: a new method for an old

problem.

Journ al of Egy ptia n

Archaeology

68: 5-15.

Petrie W. M Flinders

1901 Diospolis Pama:

he

Cem eter ies of

Abadiyeh and Hu London: Egypt

Exploration Fund.


5/68

NYAME AKUMA

No.

32 December

7989

The Mandara

rchaeological Project

1988 89

Nicholas David and Judy Sterner


University of Calgary

Calgary Alberta Canada

T N

N4

A first report on the progress of this

long-term archaeological and ethno-

archaeological project appeared in yame

kuma

29 (1987): 2-8.

Fieldwork in Cameroon

In 1989 the authors spent May through

August in the field, accompanied for part of

the time by

Dr.

David Killick, Manager of

the Archaeometry Laboratory, Peabody

Museum, Harvard University. Scott

MacEachern, a Ph.D. candidate at

Calgary, was also in the field for about a

month before being evacuated on account of

an unrelenting case of giardia combined

with amoebic dysentery.

The main emphasis of this almost

entirely ethnoarchaeological field season

was on the twin technologies of iron

metallurgy and pottery manufacture, and on

native perceptions of

these

technologies. In

particular we were interested in following

up an idea of Nigel Barley s, who had

suggested that in this

part

of Africa, social

processes such as maturation may be

perceived through a potting model . To

what extent, we wondered, were socio-

cultural, political and even natural

processes, apprehended through techno-

logical lenses? In the event we discovered

that, far from the above, indigenous

conceptions of the technologies are to a

large extent based upon physiological

models. Thus, for example, the processes of

smelting and of eating, digestion, and

excretion are seen as comparable. Such

beliefs structure the technology, including

its magical aspects, and its transmission,

and more generally underpin artifact

production.

During Killick s five weeks in the field

we arranged for the reenactment of two

traditional smelts. The first, by Dokwaza,

a Mafa, was of particular interest as its

results were markedly different from those

of a smelt by the same iron master that we

had recorded in 1986. In that year he had

produced quantities of cast iron besides

steel and iron (David et al. 19891, but in

1989 mostly low-carbon flake iron. A

portion of the product was nonetheless

successfully fined and forged into an

axhead. A second smelt, this one by a Plata

Kapa smelter living on a rocky plateau

12

m

south of Mora, had been arranged by

MacEachern before his unscheduled

departure. The Plata speak a dialect of

the language spoken by the Vame-

Mbreme , but live in Uldeme canton and are

becoming more and more integrated into

Uldeme society. Plata Kapa furnaces are

much smaller than those of the Mafa, but,

like

others in this region, work in a similar

way, air being blown down into the shaft

through a vertical tuyere. The smelt was

not entirely successful but did produce some

iron. An attempt by a Plata smith-

amongst these people the roles of smelter

and smith are distinct-to fine some older

blooms was a failure.

In addition to the two smelts we also

obtained much more information on

smelting and its products among the Mafa,

Bulahay, Mabass, Hide, Plata and Molkwo

(the last-named by courtesy of ORSTOM

colleagues), which we are beginning to be

able to synthesize into a general account of

ethnic variations in this technology. The

very different results achieved by

Dokwaza in 1986 and 1989 are a salutary

warning against the temptation to consider

traditional technologies as monolithic and

unvarying.

As a complement to traditional

recording techniques, we made video

recordings of both smelts and indeed of

much of the work carried out by David and

Killick and by Sterner. These are of

particular value in the documentation of

complex technological processes and social

occasions during which the actors are too

occupied to be able to explain what they

are doing. By showing tapes made in

1986


6/68

No

32

December 1989

to smelters, smiths, and potters, some of

whom had never seen a television

programme, we secured enthusiastic

cooperation and information that would

otherwise have taken much longer to

obtain. Such technological innovations are

of course only useful where informants

already have confidence in the

anthropologists.

In collaboration with the Cameroonian

Institute of Human Sciences we gave public

showings of the video Dokwaza: le dernier

maitre de fer africain

(David and Le Bleis

1988) to large audiences in Garoua, Maroua

and Mokolo. Dokwaza and his family were

present on two of these occasions. In answer

to questions, the iron master vigourously

expounded his claim to have received the

technology intact from his forefathers and

to have changed nothing. We believe that

the showing of the video had a

considerable impact, su ppo rting a

justifiable pride in traditional techno-

logical achievements, perhaps especially

among younger and more educated members

of non-Muslim societies whose social

position is conducive to ambivalence

towards much of traditional culture.

MacEachern worked intensively with

Plata smiths and accumulated a body

of

information on their practices and

productivity that will significantly

modify without invalidating the con-

clusions reached by David and Robertson

(in press) regarding the likely retreat of

the traditional highland school of

smithing in the face of competition from

better organized and supplied Muslim

smiths.

During a brief visit to the large Iron

Age site of Mehe

MAP 23

test excavated

in 1984, Killick found sufficient evidence on

the surface of several of the mounds to be

able to infer the practice of smelting at the

site, though not necessarily on an industrial

scale.

While the men were concentrating

their research on metallurgy, Sterner

focussed on ceramics, continuing her work

among the Bulahay of Sirak on the study of

domestic and sacred pottery and its

integration into all aspects of Sirak life.

This enquiry necessarily extends into

village history and clan origins; major

ceremonies (two of which were attended);

life histories of n g w a z l a (the smith,

smelter, potter, undertaker caste) families;

and the collection of myths about the

origins of iron, pottery, millet, and

ngwazla study of abandonment was

begun, stimulated initially by the

realisation that abandoned compounds are

among the best places to observe sacred pots

(see Sterner 1989a).

Since 1986 portions of two Sirak

quarters have been abandoned, primarily

due to lack of water. Eight compounds had

been abandoned less than six months prior

to our visit. These were sketched and

records kept on what was taken and left

behind, what will continue to

be

used there,

and what eventually will be moved. Visits

were made to several of the same families

in their new locations. The new compounds

were also sketched and more details

recorded of materials moved from one to the

other. More information on the locations

and ultimate discard of pottery, as well as

other classes of material culture, was

obtained and is contributing to further

analyses of the complex links and processes

relating objects and the culture that

generates them.

While the ritual aspects of

smelting/forging have often been described,

little has been said about the ritual

essential to the potters craft. Discussions

with Sirak and Mafa potters revealed that

potters too make offerings and sacrifices to

protect their processes and products.

Furthermore, the nature and details of

these rituals have parallels in those of

smelting/forging and the husbandry of

millet.

David also had the good fortune to be

present at parts of the Skala sla ceremony

of the Hide. This festival, of which the

high point is the release and recapture of a

bull that has been stall-fed for a year or

more, is in its various forms the central rite

among several of the peoples of this region.

We hypothesize that the release of the

bull from its stall, its recapture and return,

constitute a paradigm for human manip-

ulation of nature through the invocation

and temporary control of spirits of God,


7/68

NYAME AKUMA

No

32

December 1989

natural forces, and ancestors.

These are

induced to reside for a while in pots. The

Hide version of the bull festival is

obviously derived from that of the Mafa.

David filmed both the Hide ceremony and

a Hide funeral and burial.

David and Sterner are planning to

make a second video, tentatively entitled

Vessels of the Spirit, on the place of

pottery in the lives of the Mafa, Bulahay,

and Hide.

Laboratory nd Other Studies

During 1988 and 1989 work has

continued on various aspects of our data.

David Killick, Robert Heimann, a

ceramicist at the Alberta Research Council,

and Michael Wayman, Professor of

Metallurgy at the University of Alberta,

joined forces to study the technical aspects

of the data from Dokwaza s smelt. Their

results are given in David et al. (19891,

fully referenced below. A full technical

account is planned for Archeomaterials.

Sterner is also investigating the

temporal and spatial distribution of a

technique of pottery manufacture, largely

restricted to northern Cameroon and

Nigeria, that makes use of a concave

wooden anvil-mould and a ceramic tamper

in the production of the base and a mapr

portion of the upper part of the pot. The

Mafa and some of their neighbours use this

technique to form only the pot base,

continuing with broad coils that are

pulled up to form the upper pot body. Dale

Walde, a Ph.D. student in archaeology at

Calgary, has investigated these techniques

making use-following preliminary

experimentation by Patrick Carmichael, a

Calgary Research Associatmf methods

developed by Owen Rye that utilize

radiography to assist in the study of

techniques of manufacture. In unpublished

papers Carmichael and Walde have

confirmed Rye s interpretations of the

radiographic consequences of specific

manufacturing techniques and extended

them to the Cameroonian materials.

Walde took pots of known manufacturing

technique together with large sherds from

the test excavations at Mehe and from

surface collections obtained during the 1984

survey. He has shown that it is possible

to

discriminate radiographically between the

pounded pot technique, combined pounding

plus coils, and other techniques.

Radiography also shows considerable

promise in studies of variations in temper.

Work on both these aspects will be

continued.

Two accelerator dates have been run by

IsoTrace on samples from the Neolithic site

of Blabli (MAP 506A) tested in 1984.

This

site had previously given an unsatisfactory

set of three thermoluminescence determi-

nations ranging from 1940f 190BP to

64

f

70 BP. The new dates, both run on

ovicaprine bone fragments from the main

occupation horizon, are as follows:

TO-1127 4,390

f

220 BP

TO-1128 6,960f 200 BP

Since we have inferred that the

occupation represents a camp occupied

during a single phase, and most probably

quite briefly at some time between about

2,000 and 4,000 BP, we are uncertain how to

react to these dates.

Ph.D. dissertations by K. Gavua and

S. MacEachern (University of Calgary) and

D Lyons (Simon Fraser University) are

nearing completion.

D. Mueller (1989) and E W. Wahome

(1989) have completed their Master s

theses on MAP materials. Their abstracts

are given below.

D

Mueller Bono pn u w rt i o n in

Weat

k n t m l

Africa

This thesis focuses on the preservation

of bone collagen and the potential for stable

isotope analyses of samples from two

Northern Cameroonian sites, one Iron Age

site (MAP 523) and one Neolithic site

(MAP 506A). Two similar collagen

extraction techniques were compared, one of

which includes a step to eliminate base-

soluble humate contaminants through a

soak in a basic (NaOH) solution. This

comparison involved several variables,

including extractable yields of carbon and

nitrogen, and ratios between the stable

isotopes of these two elements.


8/68

NYAME AKUMA

No 32 December 1989

The comparison of techniques suggests for the first time established precise

that humates artificial ly inflate comparisons between the ceramics of the

extractable yields. Elimination of these study area and the nearest well-docu-

contaminants produced lower collagen mented materials from Daima in Nigerian

yields, but higher yields of carbon and

Borno.

nitrogen and more negative 613C values. It

was concluded that, while a certain amount

Acknowledaments

of collagen may be lost during humate

removal, the elimination of these

This work was supported by grants

contaminants is a necessary step for an

(410-88-0361 and 410-89-0871) awarded

acceptable analysis using stable isotopes.

David and Sterner by the Social Science

and Humanities Research Council of

Samples from the Iron Age site

Canada, and by grants from the University

generally produced higher yields of

of Calgary Vice-President s (Research)

collagen, carbon, and nitrogen when

Fund for metallurgical analyses and

compared to the Neolithic site. Within

archaeological mapping. The video was

the Iron Age site, bones from the lower sponsored by the United Steel Workers of

levels produced higher yields compared to

America, Canadian National Bureau.

bones from the upper levels. It was

concluded that past environmental Publications Papers Video and Poster

conditions (i.e., increased groundwater

Session Presented

flow), in concert with human activities,

have influenced the degree of bone

preservation at the two sites. The few

acceptable stable isotope values indicate

that humans, cattle, and sheep were

primarily feeding on C4 vegetation while

goats included some C3 vegetation in their

diets.

The potential use of stable isotope

analysis in West Africa is discussed and it

is concluded that this technique can

provide additional information to prehis-

torians. Recommendations for future

research in stable isotope analysis are put

forth.

E W Wahomm knm w

and hh o r y in

the lron Age of North Cameroon

This study is based on the Iron Age

ceramics of North Cameroon and

northeastern Nigeria dating from the first

to sixteenth century .D. This is an area

generally believed to have gone through

myriad political, social, and economic

changes during the Iron Age.

However,

I

have demonstrated an

extraordinary degree of continuity in the

ceramics of the North Cameroonian part of

the study area that contrasts with the

variability that might be expected on the

basis of the historical record for the zone

lying a short distance to the north. also

David, Nicholas, Killick, David, Heimann,

Robert, and Wayman, Michael

1989 Between bloomery and blast furnace:

Mafa iron-smelting technology in

North Cameroon. African Archae-

ological Review

7 :

18 208.

David, N. andLe Bl s, Yves

1988 Dokwaza: last of t k African iron

masters. Video, 50 min. Calgary:

University of Calgary, Department of

Communications Media. (Also

available in French under the title,

Dokwaza: le dernier ma ftre d e fer

africain

.

David, N., and MacEachern, A. Scott

1988 The Mandara Archaeological Project:

preliminary results of the 1984 season.

In D. Barreteau and H. Tourneux

(eds) Recherches comparatives et

historiques duns le bassin du Lac

Tchad pp. 51-80, Paris: Editions de

I'ORSTOM.

David, N., and Robertson, Ian

In press Competition and change in two

traditional African ironworking

industries. In P. Schmidt (ed) T h e

African lron Age past and pres ent.

Gainesville: University of Florida

Press.


9/68

NYAME AKUMA

No

32 December

1989

David, N., Sterner, Judy, andGavua Kodzo

1988 Why pots are decorated. C ur r en t

Anthropology 29 (3): 36 89.

Gavua, K.

1989 Goat skins versus wax prints: an

analysis of Mafa costume. In D. C.

Tkaczuk and B. C. Vivian (eds)

Cultures in Confl ict : Current

A r c h a e o l o g i c a l P e r s p e c t i v e s ,

pp. 293- . Calgary: Chacmool.

Lyons, D.

Deliver us from evil: protective

materials used in witchcraft and

sorcery confrontation by the Mum of

Doulo, Northern Cameroon. In D. C.

Tkaczuk and B. C. Vivian (eds)

Cultures in confl ict : Current

A r c h a e o l o g i c a l P e r s p e c t i v e s ,

pp. 297 302. Calgary: Chacmool.

Men s houses, women s spaces: the

spatial ordering of households in

Doulo, North Cameroon. In S.

MacEachern, D. Archer and R. Garvin

(eds) Households and Communities,

pp. 28-34. Calgary: Chaanool.

Time, space and gender. Paper

presented at The Archaeology of

Gend er , 22d Annual Chacmool

Conference, Calgary (November

1989).

MacEachern, A. S.

In press The people of Ngolele and the

question of Wandal a and

Montagnard origins. In Proceedings,

4e Colloque Internationale MJga-

Tchad. Paris: ORSTOM.

MacEachern, A. S., and Peacock, Sandra

1988 Ethnicity and ceramic variation

around Mayo Plata, northern

Cameroon. Paper presented at the

21st Annual

Canadian Archaeological

Association Meetings, Whistler, B.C.

Mueller, Denis. Bone preservation in West-

1989 Central Africa. M.A. thesis, University

of Calgary.

Sterner, J.

1989a Who is signalling whom? Ceramic

style, ethnicity and taphonomy among

the Sirak Bulahay. Antiquity 63:

451-

9.

1989b Sirak household ritual. In S.

MacEachern, D. Archer and R. Garvin

(eds) Households and Communities,

pp.

22-27. Calgary: Chaanool.

Sterner,J. and David, N.

1989

Gender and caste interactions in the

Mandara highlands (N. E. Nigeria and

N. Cameroon). Paper presented at

The Archa eology of Gend er,

2

2

d

Annual Chacmool Conference,

Calgary (November 1989).

Sterner

J., and MacEachern,

A.

S.

1988 Signals sent; messages received-

ceramics of north Cameroon. Poster

session, e Colloque Internationale

M4ga-Tchad, Paris.

Wahome, Ephraim Wachira.

1989 Ceramics and history in the lron Age

of North Cameroon. M.A. thesis,

University of Calgary.

Recent Archaeological

Research in the Western

Accra Plains Ghana

Leonard B. Crossland


University of Ghana

Legon-Accra Ghana

At the time of independence, a

construction boom hit the Accra

metropolitan area that has been on the

increase ever since. The city is bursting out

beyond the suburbsas housing development

increases. The Weija dam has created a

vast reservoir for irrigation. Power and

sewerage lines, as well as modern


10/68

N Y M E

AKUMA

No

32

December 7989

highways, stretch to the horizon.

Agricultural development has not only

increased the acreage placed under the

plow, but has also placed the plow deeper

into the ground. These developments have

impacted many archaeological sites in the

area. To the historian and the

archaeologist all this is appalling to

contemplate. The complexity of the

problem has gradually become apparent

and in recognition of the archaeological

values which might

be

endangered by these

construction activities, the present writer

initiated the Accra Plains Cultural

Resource Management Project in 1988. The

principal objectives of the project are:

to call attention of scholars and

scientists in government, the

Universities, the National Museums,

and traditional rulers to wanton

destruction of our cultural resources;

to identify and register all archae-

ological resources; and

to focus staff and student personnel

problems in the area.

Summary of Current Research

Ayawaw Gnat Accra)

Previous archaeological work at

Ayawaso (5 40 N,O27 W) was primarily

concerned with excavations concentrated

mainly on mounds in the Obutu quarter. As

part of the teaching programme of the

Department of Archaeology, Legon five

B.A. final year students and one M.Phi1.

student were exposed to ten days of field

training at the site in April, 1988. Survey

information, utilizing a transect interval

sampling method, with lines radiating

from a central datum ever 45 degrees,

indicated that the Ayawaso township was

very extensive. The transect survey also

confirmed the evidence of oral history

indicating that the town had five quarters

of different ages: Obutu, Obutu market,

Okaikoi, Akotia, and Ayawaso (brass bowl

shrine) quarters. The Obutu quarter

accommodated the indigenous segment of

the population and was the centre of

farming and smithing. The Okaikoi

quarter was the seat of the Ga chief. The

Ayawaso (brass bowl shrine) quarter and

the Obutu market quarters were central and

had a mixed population. The Akotia

quarter, an eighteenth-nineteenth-century

middle class enclave founded by Nee

Akotia Owosika, was the centre of trade in

European wares obtained through coastal

warehouses at Christiansborg (Danish) and

English and Dutch Accra. The surface

survey produced large quantities of local

and European artifacts which are currently

being studied by the present writer.

In March, 1989, Bredwa Mensah

completed a reconnaissance survey and

excavation of the Okaikoi quarter as a

partial fulfillment for the M.Phi1. degree

programme. In view of its location on a hill

(Kplagon) overlooking Accra from the

north, the Okaikoi quarter is currently

being threatened by urban expansion and

commercial farming activities, and the

pace of research needs to be doubled.

Wodoku

Hill

The Wodoku site (538 N,O?l W) lies

to the north of the Kotaka International

Airport. The recent installation of an

aircraft navigational device for the airport

and the construction of urban dwelling

houses have seriously impacted the site.

The potential for archaeological

research at Wodoku was recognized in the

1930s by Margaret Field, an anthropologist

who was interested in Ga social

organization. In the 1960s Oliver Davies

and Paul Ozanne completed limited surface

surveys at the site. In 1965 Dr. John

Alexander completed the first scientific

excavation, but the entire assemblage of

later Iron Age cultural material was

exported to Cambridge and his observations

have not yet been published.

A recent survey by Tawiah (1988)

utilizing the radial transect sampling

method sampling method has

demonstrated that the site stretched over

an area of approximately 21 hectares. The

Civil Aviation Authority had destroyed 5

hectares and the other 17 hectares are

threatened by new housing projects.


11/68

No

32 December 1989

Test excavations reported here took

place July 4-14,1989, under the Department

of Archaeology Field School programme

directed by the present writer.

Participants in the programme were B.A.

Final Year students: Samuel Odame,

Reverend Kwasi Quarm, Henry Nensah,

Teye Narh, and a number of student

volunteers from the

B.A.

First Year.

Funding for the project was provided by a

Department of Archaeology Research vote.

The site measured about 60

x

45 m and four

units of varying shapes and dimensions

ranging ranging from 8 m2 to 10 m2 were

excavated by 20 cm arbitrary levels.

Cultural material was largely concentrated

between 20 and

60

cmdeep, and the site was

underlain by a sterile clay layer below 80

cm. ne unit, K-69, and two trenches, T-1

and

2

revealed deep depressions cut into

the sterile clay later, possibly for clay

mining. The depressions were later filled

with domestic rubbish. The rarity of refuse

mounds at Ayawaso perhaps can be

explained by this mode of refuse disposal in

the past. Only two refuse mounds have

been recorded.

Physical observation and microscopic

examination of the potsherds showed that

the entire assemblage (n=6,010) approx. can

be divided into the following categories:

Ware

A

(97 ); Ware B-Shai Classic

(2 ); and a Miscellaneous Group (1 ).

Ware A complete vessels were rare

(n=2). The ware is characterized by paste

composed of quarts grains, iron oxide

nodules, and feldspar and occasionally

shows minute specks of mica and

hornblende derived from the underlying

Togo-Dahomeyan basement rocks. The

pottery, while not yet studied in detail, is

thin, well fired, and shows resemblances to

earthenwares from ancient Ga settlements

of Ayawaso, Opah, Medea (Salem) and

Ajenkotoku located west of Wodoku.

Diagnostic decorative motifs include

1)multiple incised and grooved lines on

the

inner surface of everted rims, 2) scoring

(striations) achieved by wiping or raking,

and

3)

modelled or appliqued fillet on the

shoulder of vessels. Vessels' shapes are

simple, globular-bodied jars with everted

and vertical rims (49 ), shallow open

bowls (43 ), and pedestalled vessels.

Ware B was found to contain quartz,

feldspar, garnet, and hornblende. The

fabric occasionally shows mica flakes

(white) but this is minimal compared to

Ladoku ware. The garnet and hornblende

are metamorphic minerals probably

derived from the metamorphic rocks of the

Shai hills and the surrounding countryside.

Garnet

seems

o have a wide distribution in

the area south of the Shai hills, and it has

been recognized in soils and clays by the

writer at Bundase near (Ladoku) and at

Tema (Dombrowski 1977: 32). The sun-ray

decorative motif identified by Anquandah

(1982: 16) as one of the diagnostic elements

of this ware is represented. Shai Classic

Ware had been identified as a dominant

import at Ladoku from the sixteenth to the

eighteenth century .D. (Anquandah,

personal communication).

A large amount of fauna was recorded,

including bones of sheep/goat, antelope,

grasxutters, domestic fowl, and cattle. An

examination of the bones revealed that

most of them were crushed. Also transverse

breakage is prevalent, possibly resulting

from narrow extraction. The mollusca and

fish remains indicate a heavy emphasis on

the exploitation of marine and lagoon

resources with less emphasis on hunting.

Carbonised remains of oil palm nuts

recovered may indicate contacts with the

Akwapim ridge along an ancient footpath

which according to oral history connected

coastal settlements of Nungua and Labadi

through Wodoku and the Sassabi plains

(Oyibi) at the foothills of the ridge.

Levels below 80 cm produced types la

and the upper levels types 2b of Ozanne's

(1965) smoking pipe type series, which are

dated relatively to ca. 1640-70. These

dates correspond fairly well with

Reindorf's estimates of 1660-80 for the

foundation and destruction of several


12/68

NYAME AKUMA

No. 32 December 1989

Nungua and Labadi settlements east of

Accra 1966. 36-44). Microscopic exami-

nation of the pipes indicate that type 2b is

of Shai origin and type l a is probably

local. Three charcoal samples were

collected from Levels

40-60

cm; 60-80 m

and 140-60 cm of Unit K-69 for C14

determination.

References

Anquandah,

J

1982

Archaeological reconnaissance and

excavation in the Shai Hills, Ghana.

N y a m k u m a

21

Dombrowski,

J

1977

Preliminary note on excavations at a

shell midden near Tema, Ghana.

N y a m k u m a

10

Ozanne, P. C.

1965

Tobacco pipes of Accra and Shai.

Bulle tin of the Insti tu te of frican

Studies

1.

Reindorf,C.

C.

1966

he history of t k Gold Coast and

sante reprint). Accra.

Tawiah, Abrefa

1988 An Archaeological Survey of the

Wodoku Site. B A Hons) Long Essay,

Protohistoric Archaeology

in the Banda Area Ghana

Ann B. Srahl

Department of Anthropology

SUNY Binghamton

Bingham on N Y 1390

1

USA

Over the course of the last four years I

have been engaged in a study of continuity

and change during the last several centuries

in the Banda area of westcentral Ghana.

The Banda Traditional Area is situated

south of the Black Volta bend and lies

immediately north of the contemporary

forest boundary Fig.

1 .

The area is

characterized by a range of high hills that

rise dramatically out of

an

otherwise low-

rolling topography. In the period from the

eighteenth century, this small, multi-

ethnic area was incorporated initially into

Asante and later British imperial

networks.

My research is directed toward an

exploration of the following related issues:

identifying the timing and processes that

linked Banda to external networks; the

temporal and regional variability of

participation in global networks; and the

implications of the colonial period for our

understanding of a traditional African

society. Archival research conducted in the

Public Records Office, London and the

Ghana National Archives, Kumasi 1987-

89) provided insight into the character of

colonial administration of Banda and has

led to the identification of changes in

ethnic identity and ethnic style during

periods of social, political, and economic

dislocation that characterized the Volta

region during the nineteenth century Stahl

1989). The archaeological component of

this research is directed toward

understanding the impact of changes at the

local level.

I

returned to Banda in the summer of

1989 to undertake archaeological

excavation at a late nineteenth-century site

in the Banda area with the goal of

identifying protohistoric patterns of

subsistence, settlement, and craft

specialization. These pat terns will

provide important comparative material

for understanding the character and timing

of subsistence changes i.e., in the adoption

of New World crops), the nature of

precolonial settlement patterns, and

assessing patterns of continuity or change in

craft specialization prior to incorporation

into the colonial market economy. Analysis

of the 1989 excavation materials is ongoing;

however, preliminary results of the

excavations are reported here.


13/68

NYAME

KUM No

32

December d989

Archival and Ethnohistoric

Background to Excavations

at Makala Kataa

Excavations were undertaken at the

site of Makala Kataa during the months of

June and July 1989. Makala Kataa

(8 8 1SWN;2 23 15 W) is a protohistoric

site located adjacent to the contemporary

village of Makala in the Banda Tradi-

tional Area of the Brong-Ahafo Region,

Ghana. site lies to the southwest of

the present settlement and is situated on a

low-lying, northeast-southwest trending

ridge. The seasonally active Selloo River,

which drains southward into the perennial

Tombe River, marks the eastern

edge

of the

Makala ridge. The Tombe, which would

have provided a reliable source of water

for the settlement at Makala Kataa, lies

approximately half a kilometer south of

the site.

Former traces of occupation are

scattered in a linear fashion along the

Makala ridge and cover an area

approximately 300

m

(northeast-

southwest) by 50 m. (northwest-southeast;

Fig.

2 .

Vegetation on the site consists of

continuous

grass

cover associated with a

variety of savanna woodland trees; there is

no evidence of recent cultivation. Baobabs,

typically found clustered on both

contemporary and ancient occupation sites,

dot the site. Wall stubs of former structures

(to

a

maximum of ca. 1 m) are visible in the

area of the site adjacent to the

contemporary village. Standing wall stubs

cease to be visible to the southwest, away

from the contemporary village (Fig.

2).

Here, the site is characterized by a series

of low mounds that appear to be the

remains of collapsed compounds. The

mounds are separated from one another by

low-lying areas.

Fig. 1 Location of t h e Banda Traditional Area.


14/68

NY ME KUM No

32

December 1989

Oral traditions collected from

Nafaanra-speaking peoples identify

Makala Kataa as a site of former Nafana

occupation. This should not be taken as an

indication that the ancient settlement of

Makala was occupied by a single ethnic-

linguistic group, for the Banda area is

characterized by considerable ethnic-

linguistic diversity today. Moreover, there

is considerable ethnic diversity that is

today subsumed within the Nafana

identify, and there have been substantial

alterations in the ethnic style or content of

Nafapa identity (Stahl 1989) over the

course of the last century. Nevertheless,

the oral traditions collected from

Nafaanra-speaking people provide insight

into the context of occupation. Traditions

collected from the Royal Family (Sielongo

Katoo) identify Makala Kataa as the

initial site of Nafana occupation following

their migration from the ancestral

settlement of Kakala in the contemporary

C6te d'Ivoire (Ameyaw 1965: 1, Stahl and

Anane 1989:

7 .

Traditions collected by

Ameyaw (1965: 8-91 relate that Makala

was abandoned during periods of upheaval

when the Nafana relocated to the north of

the Black Volta River. By the end of the

nineteenth century, the Nafana had

returned to the area south of the Black

Volta and occupied the site of Bui (York

19651, located adjacent to the contemporary

village of Bui. This was the period during

which the mounted troups of the Imam

Samori were active in the hinterland of the

Gold Coast (Ameyaw 1965: 12-13). British

troops were garrisoned at Bui (Lawra)

during the early months of 1897 (e.g.,

Maxwell 1897a, b). British and French

intervention resulted in the cessation of

slaving and warfare in the Black Volta

region during the final years of the

nineteenth century. Traditions indicate

that it was during this period that the

Nafana reestablished villages at formerly

occupied sites, including Makala.

More detailed information collected

during the 1989 field season suggests that

there are several loci of occupation

subsumed under the Nafaanra label

Makala Kataa. The site of initial

Nafana occupation is reportedly located to

the northeast of the occupation areas

described in this report, and lies adjacent to

the contemporary Kuulo village of

Dumpofie. A knowledgeable elder of

Makala village indicated that the

occupation of loci adjacent to the

contemporary village of Makala occurred

at a later period. Settlement occurred

initially in the extreme southwest of these

loci, and occupation was concentrated in the

area immediately west of the

contemporary village. The elder reported

that he was

born

and spent his youth living

in this area of Makala Kataa. The site

was abandoned when a white

representative of the colonial government,

known as the breaker of walls, visited

the area and told the inhabitants of

Makala to relocate their settlement. He

cited the close spacing of the houses as a

fire hazard, and instructed the Nafana to

reestablish their village according to a

grid plan that incorporated wide alley

ways between compounds.

his

s consistent

with archival documentation that attests

to the concerted effort by the colonial

government to establish new native

settlements that were in accord with

British planning concepts. Virtually all

contemporary settlements in the Banda

area are associated with an adjacent

k t

or site that was abandoned during the

colonial period (probably after 1930).

Archival sources place these local changes

into broader perspective. Annual

Departmental Reports issued by the

colonial government in the Gold Coast

highlight the government's concerns with

implementing changes in village layouts

that were thought to promote sanitation.

Preliminary Results of 1989

Investigations at Makala Kataa

Activities during the 1989 season

included preliminary mapping of the site,

controlled surface collection, subsurface soil

augering, and excavation. Surface

collection, augering, and excavation

concentrated on the apparently earlier

occupation areas at the southwestern end of

the site (Survey Stations 6-8; Fig. 2).

Surface collection and augering suggest that

the mounds may be differentiated into


15/68

NY ME KUM

No

32 December

1989

middens and former occupation contexts

i.e., collapsed compounds). Midden

mounds were characterized by high

densities of surface material and dark soils;

occupation mounds yielded low densities of

surface material and generally lighter

soils.

Excavation was undertaken in

contiguous 2-m2 units and focused on an

occupation mound and

an

adjacent low-lying

area separating mounds Fig. 2). total of

32 m2

8

units) of deposit was excavated to

a depth of approximately one meter, the

point at which gravelly subsoil was

encountered. ll deposits were screened( -

inch mesh) and flotation samples collected

from each 1 level.

Excavation of the mound deposits

yielded evidence of primary occupation

contexts. One unit revealed areas of

compacted gravel 3-5 cm thick) that

appear to be the remains of house floors.

Adjacent units yielded concentrations of

food-processing equipment grinding stones)

in association with hearth features.

Hearths were distinguished by the

presence of in s i b hearth stones laterite

blocks) associated with thick concen-

trations of burned sediment. Surrounding

the hearths were a number of ceramic

vessels of a variety of shapes bowls, jars)

and sizes. A number of these vessels were

intact; other ceramic concentrations appear

to

be

the remains of vessels crushed in situ.

In some instances, vessels appear to have

been stacked one on top of the other.

Concentrations of carbonized grain,

including sorghum, were recovered from

associated deposits.

Two units

8

m2 total) were excavated

in a low-lying area adjacent to the mound.

The upper 20-30 cm of these units yielded

abundant, fragmented ceramics associated

with a very dark soil suggestive of a high

organic matter content. At approximately

25 n below surface

three

concentrations of

gravelly clay were encountered that were

thought to represent wall stubs. The

o

r ~

Survey stations L d Areas of systemat~csurface collect~m

10ll

wgc ng

M Mound

Excavated areas

W Wall

Ommeter contour ~nterval ; levation arbitrary

Fig 2. The protohistoric site of Makala Kataa.

15


16/68

NY ME AKUMA

No 32 December 1989

boundaries of these concentrations were

linear (ca.

40 rn

across) and were readily

definable. Sectioning of two of the three

concentrations on the last day of

excavations, revealed that they were

burial pits, both of which were partially

excavated. Burial was an extended

inhumation of a subadult. No grave goods

were encountered. The bone was highly

friable and began to disintegrate upon

exposure. Only the cranium of Burial was

exposed. Although exposure of the burials

was cursory, the comparative information

on orientation and burial context provides

useful insight into precolonial patterns of

burial.

Preservation of faunal material was

reasonably good. The fauna has been

identified by Peter Stahl using

comparative collections at the American

Museum of Natural History in New York.

Small quantities of domestic fauna were

recovered and include Bos (cattle),

ovicaprids (sheep/goat) and guinea fowl

(Numida). A range of wild species are

represented, including several large ground-

dwelling rodents (ground squirrel, giant rat

and grass cutter), in addition to tortoise and

duiker. Observation of butchery marks is

suggestive of species-specific patterns of

butchery.

Flotation yielded large quantities of

carbonized macrobotanical material.

Samples are currently being processed, but

preliminary sorting suggests an abundance

of sorgham. This, combined with the

ubiquity of grindstones, points to the

possibility of subsistence changes since the

nineteenth century. Major crops grown in

the Banda area today include yams, and

the New World domesticates maize and

manioc (cassava). Manioc and maize are

relied upon during the rainy season when

yams are unavailable. Up until the recent

introduction of diesel-powered grinding

machines to the Banda area in 1986, maize

was processed by pounding in wooden

mortars. Sorghum is grown in limited

quantities and is used primarily in the

production of sorghum beer (pito).

Nineteenth-century occupants of Makala

Kataa clearly possessed maize, as is

evidenced by the abundance of maize cob

roulette impressions on ceramics; however,

the ubiquity of sorghum and grindstones in

nineteenth-century archaeological contexts

is suggestive of changing patterns of

reliance on indigenous vs. introduced staple

crops and changes in associated processing

technology.

Ceramics recovered from the 1989

excavations show elements of continuity

with contemporary ceramics manufactured

in

the Banda area today

see

Crossland and

Posnansky 1978). Analysis is, however,

suggestive of certain differences.

Archaeological ceramics from Makala

Kataa appear to represent a wider variety

of vessel forms than are currently made.

Combinations of decorative motifs on pots

exhibit greater heterogeneity than is the

case among contemporary ceramics. Also,

there appears to

be

some stratigraphic

differentiation in the ceramic assemblage.

Levels from the base of the mound, which

were originally thought to represent sterile

deposits, yielded ceramics made from

distinctive pastes and included painted

pottery. Although maize cob roulette occurs

in virtually all excavation levels, cord

roulette is more common in the basal levels

of all units and is virtually nonexistent in

the upper levels. Thus, ongoing analysis

promises to reveal stratigraphic/temporal

trends in ceramics.

A surprising feature of the 1989

excavations was the dearth of imported

goods. Small quantities of beads and

limited numbers of European pipes are the

only imported items identified in the

collections to date. This provides an

interesting contrast to York s excavations at

the roughly coeval site of Buli (York 1965)

in which imports were apparently

abundant. Similarly, Posnansky s excava-

tions of eighteenth-century contexts at

Begho have yielded large quantities of

imported godos, most notably pipes

(Posnansky 1979). This is suggestive of

intraregional (Bui) and temporal (Begho)

variability in the connections of the Banda

area with external exchange networks.

It is my goal to undertake further

excavations at Makala Kataa during the

summer of 1990. Additional mounds in the

vicinity of Stations 6-8 will be sampled


17/68

NY ME AKUMA

No

32

ecember 1989

and excavation will

be

expanded to include

areas adjacent to the contemporary village

that were abandoned in the early

twentieth century. These excavations will

provide a comparative sequence covering

the period of the mid to late nineteenth

century through the early twentieth

century that, when combined with the

archival and ethnohistoric data, promises

to provide insight into patterns of change

and continuity over the past two centuries.

Acknowledgments

The excavations described in this

report were funded by the Wenner-Gren

Foundation for Anthropological Research

Gr. 5133). The ethnohistoric research

conducted in 1986 was supported by the

British Academy. The support of both

agencies is gratefully acknowledged.

Excavations were licensed by Ghana

Museums and Monuments under Section

8

of

the National Museum Deaee nd the kind

assistance of the museum staff, most

notably

Dr.

I N. Debrah and Dr. Duan was

much appreciated. Logistical arrangements

were aided by Professor Peter Shinnie and

Ama Shinnie, University of Calgary, and,

as always, I am grateful for their kind

support. Analysis of fauna was undertaken

by Peter Stahl, SUNY Binghamton.

Finally, I thank Tolee Nana) Kofi Dwuru

111, Paramount Chief of the Banda

Traditional Area, for his enthusiastic

support of my research.

References

1965

Tradition of Banda. In Tradition s from

Brong-Ahafo pp. 1-15. Institute of

African Studies, Legon. Legon:

University of Ghana.

Crossland,L B., and Posnansky,M

Maxwell, Governor

1897a

Enclosure in no. 191, Instructions for

the guidance of officers employed in

the Gold Coast hinterland, April 7,

1897. CO 879/48, no. 529,

Further

Correspondence in the Bend of the

N i g e r . London: Public Records

Office.

18971, Governor Maxwell to Mr.

Chamberlain, August 16,1897, no. 131

in CO 8798/50, no. 538.

F u r t h e r

Correspondence in the Bend of the

N i g e r .

London: Public Records

Office.

Posnansky,M

1979 Excavations

at

Begho, Ghana, 1979.

Nyame Akuma

15:

23-27.

Stahl,

Ann B

1989

Ethnic styles and ethnic boundaries: a

diachronic case study from west

central Ghana. Submitted to

Man.

Stahl, Ann B., and Anane, James.

1989 Fam ily histories from the Bandat

Traditional Area Brong-Ahafo Region

G h a n a . Report on file, Institute of

African Studies, Legon and Ghana

National Museum.

York,

R

N

1965 Excavations at

Bui: a

preliminary

report. West African Archaeological

Newsletter

3:

18-21.

1978

Pottery, people and trade at Begho. In

The Spatial Organisa tion of Cu ltu re

ed by I Hodder, pp. 77-89. London:

Duckworth.


18/68

NYAME AKUMA

No. 32

December 1989

Preliminary Report on

Excavations at Okai Koi Hill

Site Ayawaso), Western

Accra Plains of Ghana

Brenda-Mensah Yaw


University of Ghana

Legon

20

Ghana

From February-December, 1988, an

archaeological research was carried out at

Okai Koi Hill Site (Ayawaso) to provide a

corpus of Ga cultural material for the

reconstruction of the early history of the

Ga, who today inhabit the western Accra

Plains of Ghana.

Ayawaso (5 4O'NIO'l8'W) was the

settlement early European writers referred

to as Great Accra, the traditional capital

of the Ga located 21 km north of modem

Accra (Wilks 1957; Berry 1958). This

settlement was destroyed by the Akwamu

in 1677.

Some amount of archaeological

research has been carried out in the area

before (e.g., Ozanne 1962, Anquandah 1978).

These researches were, however, limited in

scope. All the excavations were restricted

to one particular rubbish mound located in

the Obutu (Awutu) Quarter and described

as the largest on the site (Ozanne 1962),

Anquandah 1982). The present knowledge

of the culture history of the area is thus

very limited. This study employed

archaeological and nonarchaeological tools

to provide data that will allow a broader

and more detailed understanding of the

history of the western Accra Plains during

the Iron Age.

Problem Focus

The objective of the research was to

reconstruct the early history of the Ga as

revealed by their material culture. Within

this context, the research set itself to

resolve some specific questions:

Was there a pre-sixteenth-century

settlement and, if any, what does the

archaeological data reveal about it?

Did the Ayawaso settlement end

abruptly or declined gradually after

the late seventeenth-century Akwamu

invasion?

What was the economic pattern of the

peoples who inhabited the research

area and in what ways was it related

to the neighboring ethnic groups?

Research Approach and Results

The nature of the archaeological

problems outlined above demanded that an

interdisciplinary approach be adopted.

The research therefore looked beyond

archaeology for help from sources such as

ethnography, oral traditions, ethnobotany,

and linguistics to provide information that

will assist in the analysis and

interpretation of the archaeological data.

The research was carried out in three

phases: (February-April) involved data

gathering on the human and economic

geography of the research area. The area's

contemporary traditional modes of life-

farming, fishing, hunting and gathering,

and local craft industries-were studied.

This period was also used to document

oral traditions on Ga origins and the Ga-

Akwamu conflict of the late seventeenth

century A D that led to the defeat of the

Ga and their subsequent movement to their

present settlement of Accra on the coast.

During (May), an intensive ground

reconnaissance survey of the Ayawaso area

was carried out. This was done to gain a

general impression of the research area.

With a team of six undergraduate students,

a foot survey using the Spoked Wheel

Model of radial transects was carried out to

locate, identify, and record the distribution

of archaeological sites against the natural

geographical and environmental

background and to surface collect artifacts

within sample units along the transect

lines.

The survey yielded fruitful results.

Four sites were discovered within the

Ayawaso settlement: Okai Koi Hill, Obutu

(Awatu) Quarter, Nii Akotia, and

Ayawaso Central. These sites together

constitute the entire Ayawaso settlement.


19/68

N Y A M E A K U M A N o 32 December 1989

The results have been confirmed by the

elders of Ga Mashie (modern Accra).

During (November-December), archae-

ological excavations were conducted at

Okai Koi Hill Site (Ayawaso) by the

writer, assisted by a team from the

Department of Archaeology. The site is

straddled on top of a hill named after one

of the early kings of Ga. The hill, 118 m

a.s.1. is part of a discontinuous chain of

hills that mark the southern limits of the

Akuapem Range. The Accra-Nsawam

asphalt-surfaced road and the railway

line lie at the foot of this hill.

Four pits each 3 x 3 m were excavated.

One of the pits dug on a rubbish mound

yielded the greatest number of finds. This

pit revealed a stratigraphic series from 0 to

540 cm. The three others dug on level

ground revealed almost uniform profiles

from to 180

cm

Over 10,000 potsherds were recovered

from the dig. Classification is still going

on, but the pottery is not different from

what Ozanne and Anquandah obtained

from their excavations at the Obutu

(Awutu) Quarter. There are two basic

pottery forms-bowls and jars. The bowls

are of shallow and deep forms with either

inturned mouths or with everted rims. The

jars also have globular bodies with wide

and very sharply everted rims. They are

hard and fired to either light-brown or

black colours. Most of the vessels have

cylindrical pedestals with everted foot

rings.

Over fifty pieces of locally made

smoking pipes turned up in the excavations.

Ozanne (1962) has proposed a relative

evolutionary typological sequence of

smoking pipes from the area. He

distinguished three types: Type I, 11, and

I11 with two varieties of each and a range

of absolute dates deduced from the

association of the pipes with datable

European sgraffito plate. When a detailed

study of thee pipes recovered from the

excavations is made, it will be possible to

s whether Ozanne s typology is valid.

large number of invertebrate shells

representing two types of land snails and

over a dozen species of aquatic molluscs

were obtained. The land snails were

Achatina achitana and Archachatina. Of

the aquatic shell-fish, the following

appeared to be the most commonly

exploited: Arca senilis, Pecten, Do nax

acutangulus, Tympanctus fuscata, and

Natica maronchiensis.

Bones of wild and domestic animals

(mainly bovida) were also recovered from

the excavations. A detailed report is

expected from the Zoology Department,

Legon.

A number of locally made beads were

excavated, mainly made of two types of

rock (quartz and bauxite) and marine

shells. Iron and copper/brass objects in

fairly good condition were also recovered

during the dig. A small number of

intricately worked ivory objects were found,

including a small comb, a broken knife

handle, two broken bangles, and what

seems to be a broken comb handle. One

phalange bone and four animal teeth

(incisors) were found with holes bored in

them. They had been polished to various

shapes. These might have been worn as

pendants.

A number of European imports attesting

to the trans-Atlantic sea trade with the

Accra coastland turned up in the

excavations. They include three pieces of

smoking pipes, ten pieces of wheel-made

pottery, and over twenty beads.

cknowledgment

This research was financed by the

Council for the Development of Economic

and Social Research in Africa

(C.O.D.E.S.R.I.A.), Dakar, Senegal, to

whom I would like to extend my sincere

thanks.

References

Anquandah J

R

1978

The Accra Plains archaeological and

historical project report.

Ny

a m e

Akuma 12: 24 27.

1982

R e d i s c o v e r i n g G h a n a s P a s t .

Longman.


20/68

N Y M E K U M No 32 December 1989

Berry J

1958 The Place Names of Ghana.

London.

Ozanne P

1962

Notes on the early historical

archaeology of Accra.

Transactions of

the Historical Socie ty of Ghana

6:

51

70.

Wilks I

1957

The rise of the Akwamu Empire.

Transac tions of the Historical Society

of Ghana 3: 101.

New Furnace Type from

the North of Igboland

Livinus Chikpandu Ekechukwu


University of Nigeria

Nsukka Nigeria

Archaeological survey work was

carried out at Idoha in Igbo-Etiti Local

Government Area of Anambra State of

Nigeria. This represents a continuation of

an ongoing research aimed at throwing

some light on the nature of early iron

technology in Igboland. During the survey

of the town, two types of iron-smelting sites

were found. The first type consisted of sites

located in circular depressions, each having

a crescentic mound of slag heaped above the

depression. This is quite similar to what

was observed at Umundu, an iron-smelting

site located about 10 km northeast of

Nsukka in Anambra State of Nigeria

Anozie 1979). Fragments of baked clay,

probably furnace wall debris, were also

found in and above the circular depressions

at Idoha. The second type, represented by

only one site, was located on a flat terrain.

On this site, two similar structures of baked

clay also suspected to be furnace remains

were observed. When this site was later

excavated, it revealed the presence of a

dome furnace, a rather uncommon phe-

nomenon in this part of Nigeria.

The furnace was built with prepared

clay, the type usually prepared in the

traditional way for building houses and

compound walls in most parts of Igboland.

It is dome-shaped and has a diameter of

about 100 cm close to its base and a wall

thickness of about 10 cm. It is estimated

that about one-third of the furnace

chamber is located inside the ground, and

the chamber has a clay lining. The furnace

has two funnel-shaped openings located

near the base; each has a maximum

diameter of 55 cm. These openings were

probably used for admitting tuyeres into

the furnace or for off-loading the contents of

the furnace after smelting. They could

equally have served as channels for

introducing fire into the chamber before

smelting. The chamber contained ash up to

about 50

m

deep. Some unburnt ore mixed

with slag was found adhering to the

furnace wall, and one such ore/slag block

weighed 2.70 kg.

The furnace has a platform on the

outside made from prepared clay. This

platform, or fortification, about 40 cm high,

could have been made after the furnace was

built and used in consolidating it at the

base. This practice of consolidating mud

walls at the base is very common among

Igbo traditional house builders, who use it

to protect the base of houses from damage

by rainwater dripping from the roof. Since

the upper part of the furnace above the

ground surface had collapsed, a

reconstruction of its original height was

attempted, producing a height of about

1.2

m.

None of the elders still alive in Idoha

today has any idea when smelting started

and ended in the area. Oral traditions

gathered from them, however, claim that

those who engaged in the practice of iron

smelting were their parents and

grandparents, who simply narrated to them

how smelting was carried out in the past.

Since none of the elders witnessed smelting,

it seems the practice was pre-European in

the area. Some charcoal samples from the

excavation have been sent for C14 dating in

order to help determine the age of the site.

By this discovery at Idoha, it seems that at

least three forms of iron-smelting furnaces


21/68

NYAME

AKUMA

No 32

December 7989

were traditionally used in parts of

Igboland: the pit, the dome, and the shaft

furnace.

References

Anozie,

F.

N

1979

Early iron technology in Igboland:

Lejja and Umundu. West frican

Journal of rchaeology

9: 11934.

Traditional Iron-Smelting

Methods by the Berom of

Plateau State Nigeria

J. F. Jemkur

Centre for Nigerian Cultural Studies

Ahmadu Bello Universlty

Zaria Nigeria

Presently, this writer is conducting

field research on the subject of traditional

methods of iron smelting in parts of the

Northern States of Nigeria. Already some

data on the subject have been collected from

parts of Plateau, Bauchi, Sokoto, Katsina,

Borno, and Kaduna States. It is already

apparent from these preliminary field

investigations that the types of smelting

furnaces, and therefore the methods of

smelting, differ considerably from area to

area. It is hoped that these differences

may

help in tracing the origins and former

distribution of the many different peoples

of Nigeria. It would be of great interest,

therefore, if as a first step we could

assemble records of all the sites where

furnaces are known to exist. In certain areas

there are some old men still living who

have done the smelting themselves and

have a near-accurate memory of the

process. It is obviously a matter of urgency

that their knowledge should be recorded.

Iron has been made and used in Africa

for a very long time. Archaeologists have

shown that it was coming into general use in

Egypt some 2,500 years ago and by 100A D

it was evidenced in many other parts of the

continent, including Nigeria (Pearce 1960,

Fagg 1969, Shaw 1969, Tyelecote 1975).

Nowadays iron is so plentiful that it

requires an effort of the imagination to

think how people lived before there was

any iron for their tools and weapons. What

is new of course is the virtually unlimited

supply of the metal which comes to Africa

from Europe. Until some eighty years ago,

the Nigerian people, using incredibly

uneconomic processes, toiled to produce

almost all the iron which they needed.

Certain ethnic groups used to specialise in

smelting the local ores to make iron hoes,

knives, spears, and useful tools and

ornaments. The locally smelted iron was

very tough and it is claimed that a hoe

made from it would last at least three

times as long as one made with ordinary

European mass-produced iron; and, if it

were not such a labour to do the smelting,

hoes would still be forged from locally

produced metal (Magajin Dodo Nok,

personal communication).

The

Berom

The Berom are the largest single ethnic

group on the Jos Plateau, occupying some

eighty villages within the present Jos and

Barakin Ladi Local Government Areas

with an estimated population of about

687,828 people. This estimate is based on

the 1963 census at the growth rate of 2.5

per annum. They live in an area of about

3,652 km2 centered on the intersection of

latitude 10 and longitude 972. They can

be divided into three geographical regions:

East Fan, Forom, Heipang, Du, Ropp

West Gyel, Kuru, Vwang, Riyom

South Bachit, Gashish

The various Berom groups have

different traditions of origin. According to

Gunn (1952) they came from the forest

country somewhere to the south, perhaps

from the area once referred to as the Jema's

Federation (now known as Southern Kaduna

State), moved north on to the Plateau at

Ashono in Bachit village area before

founding Riyom from where they spread

over the southern half of the Jos Plateau

Platform in three groups. It is said that one

of the groups settled on the Naraguta Hills


22/68

NYAME

KUM

No

32

December

7989

where they joined others, probably the

Jarawa. The second branch, the Anaguta

group, moved towards Kwoi, while the

third became the Gana Wuri.

Davies is of the opinion that the Berom

are composed of peoples from the northeast,

merged with a later immigration of Jukun

from the south (Davies 1949). On the other

hand, some of the Berom themselves (from

Forom) argued that they originated just

there on the Forom Mountain among the

rocks. It should be stated, however, that

all traces of the migration process of the

Berom are based on hypothesis and

mythology as not much investigation has

been made on the question of their origins

and migration into the Plateau Highlands.

Berom Iron Smelting Methods

The Berom were once great smelters and

workers of iron. It is said that the largest

trade or industry the Berom have probably

ever known was iron smelting, which is now

extinct though the holes for smelting

furnaces can still be seen over most of the

Berom area today.

The origin of iron smelting among the

Berom is not known. The Nok culture, best

known for its remarkable terracotta

figurines and iron smelting from about 500

B C

to 600 A D (Fagg 1959, Jemkur 1986) is

situated directly southwest of the present

Berom land. The Berom could have been

part of, or come into contact with, the Nok

culture. It is interesting to note that recent

investigations on traditional iron smelting

by this writer on the Berom on the Jos

Plateau and the Jaba and Ikulu people

within the Nok culture area show that the

ironworking apparatus and techniques are

similar in both areas. For example, the use

of pulsaters, skin bellows, and charcoal are

common to both groups. In addition, both

usually have their furnaces built on stream

banks (Jemkur, 1989).

Until the 1940s, when cheaper and

better imported European iron replaced the

locally smelted iron, the Berom had a

flourishing iron-smelting industry. It was

the locally produced metal ingots (wrought

iron) that were worked by the village

blacksmith into hand hoes for farming,

arrowheads, anklets, bracelets, bells that

were attached to horse harness, and

smaller bells for adorning girls. The ingots

were also traded with neighbours to the

south.

Available information describing the

processes of Berom iron smelting are scanty.

The following was obtained by the writer

from interviews with some Berom old men

from Forum and Du areas who claimed to

have participated in the smelt as youths

and could vividly remember the process.

Iron-bearing rocks were first collected.

Trees were then cut for the making of

charcoal and the furnaces were then built

along stream banks. The furnace was

usually given a preliminary heating by

burning charcoal under blast. When the

furnace had been sufficiently preheated,

charcoal and hematite ore were fed into

the open mouth at the top of the structure.

The run that produced the regulus of mixed

nodules of ore and slag started at daybreak

(preliminary heating) and continued until

about 3

P M

when the first runoff took

place. After partial cooling this regulus

was broken to remove as much slag as

possible, then remelting was camed out to

produce a cleaner and larger ingot of metal.

If necessary, this was continued until an

ingot clean enough for blacksmithing was

obtained. The operations sometimes lasted

well into the night. When the fumace was

tapped the runoff flowed into moulds

which were prepared near the front of the

furnace in a bed of clay laid there for the

purpose. The bellows consisted of two

bladders, made of goat skin which had

been removed whole (without any split)

from the goat, with straight wooden

handles fixed to the tops and wooden tubes

reaching from the bladder bottoms into the

mouth of the clayed oven channel built at

the side of the gully and leading into the

furnace, a small distance above the lower

blow hole. Blast was created by the

handles being pumped up and down at a

very rapid rate (4-5 strokes per second) by

one man.

As

one manipulator got tired, he

was relieved by another, his relief taking

over the movement in such a way that

there was little or no interference with the


23/68

NY ME

KUM

No

32 December 7989

rhythm, thus avoiding any serious drop in

the airflow (Fig. 1).

One of my informants (Choji Pam) gave

the following description of how the Berom

obtained iron:

To be able to produce iron for

making a hoe, one must be physically

strong, for it entails a lot of hard work.

Trees must be cut down and burnt to

make charcoal, which is a form of

carbon, for the production of iron is a

reduction process, with charcoal as the

reducing agent.

When enough charcoal has been

obtained, a temporary blacksmith s

workshop, called a

fwam (furnace), is

made, preferably near a stream.

When this has been done, the iron ore,

which is a gravel in the form of a hard

rock, is dug and broken into small

pieces. The iron ore can be found

today in the neighbourhood of Ray

field, Vom and Jos Aerodrome. Some

of the charcoal is ignited in the fwam.

A little of the ore is added to this

glowing charcoal. After an interval of

about ten minutes, more charcoal is

added, and more iron ore. This process

is continued from morning till night,

when it is left to cool.

The following morning the

resultant product, in the form of little

specks of iron stuck here and there in

the lump of charcoal, is removed from

the fwam ground, and winnowed.

What is left then is specially treated

with mud and straw and moulded into

spheres. When these spheres are dry,

they are taken to a proper blacksmith s

workshop and put again into glowing

charcoal. When white-hot the spheres

are taken ou t of the furnace and

beaten into the shape required.

This is how iron was made by my

people, the Berom, long before the

Europeans ever thought of coming to

Africa for legitimate trade.

From the foregoing account it is clear

that there are still some old men living

who had learnt the art of iron smelting as

youths and still remembe

nyame akuma issue 032

Documents